MRI-based fat double bond mapping

1. A method for using magnetic resonance imaging (MRI) to characterize fat, comprising:

acquiring a magnetic resonance (MR) image that includes MR data from a target;

determining fat characterization parameters based on the acquired MR data, wherein the fat characterization parameters include an amount of fat, an amount of water, an initial phase, a field inhomogeneity, a T2*, and at least one fat spectrum parameter, and wherein the at least one fat spectrum parameter includes a number of double bonds (NDB), a number of double-double bonds (NDDB), or a chain length (CL) of fat from the target in the MR data; and

using the determined fat characterization parameters to produce a relationship between regions of fat, regions of water, or regions of fat and water in the MR image,

wherein the determining fat characterization parameters comprises:

selecting initial values of the fat characterization parameters;

estimating values of the fat characterization parameters by iteratively minimizing error between the fat characterization parameters and the acquired MR data;

generating an MR signal by simulation using the estimated values of the fat characterization parameters; and

matching the simulated MR signal with an actual MR signal based on the acquired MR data,

wherein the determining the fat characterization parameters is in accordance with:

s ( t )=[ w ·α( t )+ f ·β( t )]·exp( i φ)·exp(− R 2* t )·exp( iΔB 0 t )

α( t )= a water ·exp( iωd water t )

β( t )=Σ a j ( NDB )·exp( iωd j t ),

where: a water represents the number of protons in water, a j represents the number of protons in fat, d water represents chemical shifts in water, d j represents chemical shifts in fat, ΔB 0 represents the field inhomogeneity, R2* represents transverse decay (T2*), φ represents the initial phase, w represents the amount of water, and f represents the amount of fat.

2. The method of claim 1 , wherein the CL is determined from NDB by a linear approximation.

3. The method of claim 2 , wherein the linear approximation is substantially CL=16.8+0.24·NDB.

4. The method of claim 1 , wherein the NDDB is determined from NDB by a non-linear approximation.

5. The method of claim 4 , wherein the non-linear approximation is substantially NDDB=0.092·NDB 2 .

6. The method of claim 1 , wherein the amount of water and the amount of fat are estimated by selecting an initial phase of water and an initial phase of fat to be equal at an echo-time of zero.

7. The method of claim 1 , wherein the relationship further includes regions of triglycerides.

8. The method of claim 1 , wherein the relationship further includes a number of double bonds.

9. The method of claim 1 , wherein the determined fat characterization parameters describe a composition of fat in the target.

10. The method of claim 9 , wherein the composition of fat comprises at least six known peak areas, whereby peak area 1 and peak area 2 are determined based on the peak area 3 , the peak area 4 , the peak area 5 , and the peak area 6 .

11. The method of claim 1 , further comprising:

using the relationship to characterize at least one of a composition of fat, a quantity of fat, a localization of fat, or a change in spatial distribution of fat.

12. The method of claim 1 , wherein the target includes a human subject.

13. An MRI system to characterize fat, comprising:

an MRI machine that acquires at least one MR image from a target; and

a processing unit configured to:

command the MRI machine to acquire at least one MR image that includes MR data from the target;

determine fat characterization parameters based on the acquired MR data by:

selecting initial values of the fat characterization parameters,

estimating values of the fat characterization parameters by iteratively minimizing error between the fat characterization parameters and the acquired MR data,

generating an MR signal by simulation using the estimated values of the fat characterization parameters, and

matching the simulated MR signal with an actual MR signal based on the acquired MR data; and

use the determined fat characterization parameters to produce a relationship between regions of fat, regions of water, or regions of fat and water in the MR image,

wherein the fat characterization parameters include an amount of fat, an amount of water, an initial phase, a field inhomogeneity, a T2*, and at least one fat spectrum parameter, and wherein the at least one fat spectrum parameter includes a number of double bonds (NDB), a number of double-double bonds (NDDB), or a chain length (CL) of fat from the target in the MR data;

wherein the determining the fat characterization parameters is in accordance with:

s ( t )=[ w ·α( t )+ f ·β( t )]·exp( i φ)·exp(− R 2* t )·exp( iΔB 0 t )

α( t )= a water ·exp( iωd water t )

β( t )=Σ a j ( NDB )·exp( iωd j t ),

 where: a water represents the number of protons in water, a j represents the number of protons in fat, d water represents chemical shifts in water, d j represents chemical shifts in fat, ΔB 0 represents the field inhomogeneity, R2* represents transverse decay (T2*), φ represents the initial phase, w represents the amount of water, and f represents the amount of fat.

14. The MRI system of claim 13 , wherein the processing unit is configured to use the relationship to characterize at least one of a composition of fat, a quantity of fat, a localization of fat, or a change in spatial distribution of fat.

15. The MRI system of claim 13 , wherein the CL is determined from NDB by a linear approximation, wherein the linear approximation is substantially CL=16.8+0.24·NDB.

16. The MRI system of claim 13 , wherein the NDDB is determined from NDB by a non-linear approximation, wherein the non-linear approximation is substantially NDDB=0.092·NDB 2 .

17. The MRI system of claim 13 , wherein the target includes a human subject.

18. A computer program product comprising a nonvolatile computer-readable storage medium having instructions stored thereon, the instructions comprising:

code for acquiring an MR image that includes MR data from a target;

code for determining fat characterization parameters based on the acquired MR data, wherein the fat characterization parameters include an amount of fat, an amount of water, an initial phase, a field inhomogeneity, a T2*, and at least one fat spectrum parameter, and wherein the at least one fat spectrum parameter includes a number of double bonds (NDB), a number of double-double bonds (NDDB), or a chain length (CL) of fat from the target in the MR data; and

code for using the determined fat characterization parameters to produce a relationship between regions of fat, regions of water, or regions of fat and water in the MR image,

wherein the code for determining fat characterization parameters comprises:

code for selecting initial values of the fat characterization parameters,

code for estimating values of the fat characterization parameters by iteratively minimizing error between the fat characterization parameters and the acquired MR data,

code for generating an MR signal by simulation using the estimated values of the fat characterization parameters, and

code for matching the simulated MR signal with an actual MR signal based on the acquired MR data,

wherein the determining the fat characterization parameters is in accordance with:

s ( t )=[ w ·α( t )+ f ·β( t )]·exp( i φ)·exp(− R 2* t )·exp( iΔB 0 t )

α( t )= a water ·exp( iωd water t )

β( t )=Σ a j ( NDB )·exp( iωd j t ),

 where: a water represents the number of protons in water, a j represents the number of protons in fat, d water represents chemical shifts in water, d j represents chemical shifts in fat, ΔB 0 represents the field inhomogeneity, R2* represents transverse decay (T2*), φ represents the initial phase, w represents the amount of water, and f represents the amount of fat.

19. The computer program product of claim 18 , further comprising code for using the relationship to characterize at least one of a composition of fat, a quantity of fat, a localization of fat, or a change in spatial distribution of fat.

20. The computer program product of claim 18 , wherein the CL is determined from NDB by a linear approximation, wherein the linear approximation is substantially CL=16.8+0.24·NDB.

21. The computer program product of claim 18 , wherein the NDDB is determined from NDB by a non-linear approximation, wherein the non-linear approximation is substantially NDDB=0.092·NDB 2 .

22. The computer program product of claim 18 , wherein the target includes a human subject.

23. An MRI method, comprising:

modeling a MRI-generated fat spectrum from an MR image of a target using at least one of three parameters including a CL, an NDB, or an NDDB, including:

determining fat characterization parameters based on the MR image, wherein the fat characterization parameters include an amount of fat, an amount of water, an initial phase, a field inhomogeneity, a T2*, and at least one fat spectrum parameter, and wherein the at least one fat spectrum parameter includes a number of double bonds (NDB), a number of double-double bonds (NDDB), or a chain length (CL) of fat from the target in the MR data; and

using the determined fat characterization parameters to produce a relationship between regions of fat, regions of water, or regions of fat and water in the MR image,

wherein the determining fat characterization parameters comprises:

selecting initial values of the fat characterization parameters;

estimating values of the fat characterization parameters by iteratively minimizing error between the fat characterization parameters and the acquired MR data;

generating an MR signal by simulation using the estimated values of the fat characterization parameters; and

matching the simulated MR signal with an actual MR signal based on the acquired MR data,

wherein the determining the fat characterization parameters is in accordance with:

s ( t )=[ w ·α( t )+ f ·β( t )]·exp( i φ)·exp(− R 2* t )·exp( iΔB 0 t )

α( t )= a water ·exp( iωd water t )

β( t )=Σ a j ( NDB )·exp( iωd j t ),

 where: a water represents the number of protons in water, a j represents the number of protons in fat, d water represents chemical shifts in water, d j represents chemical shifts in fat, ΔB 0 represents the field inhomogeneity, R2* represents transverse decay (T2*), φ represents the initial phase, w represents the amount of water, and f represents the amount of fat.

24. The method of claim 23 , wherein the modeling comprises measuring a number of double bonds based on the MR image.

25. The method of claim 23 , further comprising:

using the relationship to characterize at least one of a composition of fat, a quantity of fat, a localization of fat, or a change in spatial distribution of fat.